Systems and methods for controlling production and distribution of consumable items based on their chemical profiles

Abstract

Embodiments of the present invention relate to a platform for controlling production and distribution of consumable items based on machine learning processes derived from chemical profiles of the consumable items.

We claim: 
     
       1. A computer-implemented method, the method comprising:
 receiving chemistry panels associated with at least one consumable item, each chemistry panel being associated with and identifying chemical attributes of one of the at least one consumable items, the chemical attributes represented by associated quantitative chemical data values; 
 building a plurality of different executable machine learning models, each of the plurality of machine learning models being trained on a subset of the received chemistry panels associated with the at least one consumable items; 
 storing the plurality of trained machine learning models in a storage memory; 
 receiving an indication of a selection of one or more of the at least one consumable items, the indication of the selection representing an interest of a user; 
 executing of at least one of the plurality of trained machine learning models using the received indication of the selection of one or more of the at least one consumable items and at least a second subset of the received chemistry panels associated with the at least one consumable items as inputs to the at least one trained machine learning model execution to generate an output derived from the second subset of the received chemistry panels, wherein the plurality of different executable machine learning models includes at least one of a Gaussian Mixture Model (GMM) to generate an algorithmic consumable item and the at least one of the plurality of trained machine learning models execute d is the GMM, wherein an input thereto is the second subset of the received chemistry panels and the GMM is to generate a new algorithmic consumable item that fits within a distribution of chemistry determined for the second subset of the received chemistry panels; and transmitting a result set based on the generated output to the user. 
 
     
     
       2. The method of  claim 1 , wherein the at least one consumable item is wine, including bottles of wine distributed via at least one of a website, a physical store, and a wine club. 
     
     
       3. The method of  claim 1 , wherein each consumable item is identified by a stock keeping unit (SKU) and each of the received chemistry panels is associated with one of the SKUs. 
     
     
       4. The method of  claim 1 , wherein the received chemistry panels associated with the at least one consumable items are subjected to a standardization process prior to being used in building the plurality of different executable machine learning models. 
     
     
       5. The method of  claim 1 , wherein the plurality of different executable machine learning models further includes at least one of a Dynamic Time Warping model to determine how different the chemistry panels for the consumable items are from each other, a user level model to provide a recommendation to the user, and a deep learning model to perform at least one function based on the chemistry panels. 
     
     
       6. The method of  claim 1 , wherein the input to the GMM is a subset of the second subset of the received chemistry panels determined based on a user input criteria. 
     
     
       7. A system comprising:
 a processor; 
 a storage device; and 
 a memory in communication with the processor, the memory storing program instructions, the processor operative with the program instructions to perform the operations of: 
 receiving chemistry panels associated with at least one consumable item, each chemistry panel being associated with and identifying chemical attributes of one of the at least one consumable items, the chemical attributes represented by associated quantitative chemical data values; 
 building a plurality of different executable machine learning models, each of the plurality of machine learning models being trained on a subset of the received chemistry panels associated with the at least one consumable items, the indication of the selection representing an interest of a user; 
 storing the plurality of trained machine learning models in the storage device; 
 receiving an indication of one or more of the at least one consumable items; 
 executing at least one of the plurality of trained machine learning models using the received indication of one or more of the at least one consumable items and at least a second subset of the received chemistry panels associated with the at least one consumable items as inputs to the at least one trained machine learning model execution to generate an output derived from the second subset of the received chemistry panels, wherein the plurality of different executable machine learning models includes at least one of a Gaussian Mixture Model (GMM) to generate an algorithmic consumable item and the at least one of the plurality of trained machine learning models executed is the GMM, wherein an input thereto is the second subset of the received chemistry panels and the GMM is to generate a new algorithmic consumable item that fits within a distribution of chemistry determined for the second subset of the received chemistry panels; and transmitting a result set based on the generated output to a user. 
 
     
     
       8. The system of  claim 7 , wherein the at least one consumable item is wine, including bottles of wine distributed via at least one of a website, a physical store, and a wine club. 
     
     
       9. The system of  claim 7 , wherein each consumable item is identified by a stock keeping unit (SKU) and each of the chemistry panels is associated with one of the SKUs. 
     
     
       10. The system of  claim 7 , wherein the received chemistry panels associated with the at least one consumable hems are subjected to a standardization process prior to being used in building the plurality of different executable machine learning models. 
     
     
       11. The system of  claim 7 , wherein the plurality of different executable machine learning models further includes at least one of a Dynamic Time Warping model to determine how different the panels of chemicals for the consumable items are from each other, a user level model to provide a recommendation to the user, and a deep learning model to perform at least one function based on the of the received chemistry panels. 
     
     
       12. The method of  claim 7 , wherein the input to the GMM is a subset of the second subset of the received chemistry panels determined based on a user input criteria.